Back

Tumour-stroma signaling drives a reversible program with neural-like features in cancer-associated fibroblasts

Ibanez-Rios, M. I.; Batardiere, M.; Dianati, E.; Serwe, G.; Kanagasingam, S.; Archambault-Marsan, A.; Thibault, P.; Therrien, M.; Trinh, V. Q.-H.; Knapp, D. J.

2026-10-09 cancer biology
10.64898/2026.10.07.757165 bioRxiv
Show abstract

Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive cancers, in part due to a dense and highly plastic tumor microenvironment dominated by cancer-associated fibroblasts (CAFs). In previous work, we identified a CAF population expressing a neural-associated NES+SOX4+NOTCH1+ program that increased with disease progression. The signals driving this state and the mechanisms regulating it remained unclear. Here, we investigated how tumor-derived signals activate this program and whether it represents a stable fibroblast identity or a reversible state. Using spatial transcriptomic data spanning normal pancreas to invasive carcinoma, we identified EGF, FGF2 and BDNF as candidate tumor-to-fibroblast signals associated with disease progression. Functional experiments in patient-derived CAFs and human fibroblast cell lines showed that EGF and FGF2 are sufficient to induce the NES+SOX4+NOTCH1+ program. Growth factor withdrawal reversed the NES+SOX4+NOTCH1+ program at both the protein and transcriptional levels, indicating that this state is dynamic and environment dependent. Single-cell transcriptomic analysis further revealed a multi-phase response involving early loss of fibroblast-associated features, transient activation of neural-associated genes, and a later stress-associated state. Regulatory analyses identified ETS-family transcription factors as key regulators of the program, while phosphoproteomic profiling and inhibitor studies showed that MAPK/ERK signaling plays a central role in its transcriptional activation. Finally, pan-cancer single-cell analysis showed that this state is enriched in CAFs and increases from normal to tumor tissue across multiple cancer types. Together, these findings define a reversible, growth factor-driven CAF program and identify signaling mechanisms that may provide opportunities to selectively modulate tumor-supportive fibroblast states.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.